Management And Accounting Web

Ruhl, J. M. 1995. Activity-based variance analysis. Journal of Cost Management (Winter): 38-47.

Summary by Lee Salemi
Master of Accountancy Program
University of South Florida, Fall 2004

ABC Main Page | ABM Main Page | Variance Analysis Main Page

The purpose of this article is to present activity-based variance analysis that can be used by managers to increase productivity and reduce costs. Ruhl argues that traditional variance analysis reflects cost systems designed for external reporting purposes and is not very useful to managers. The author gives credit to Robin Cooper for many of the ideas presented in the article.

Omega Technology

To demonstrate the differences between traditional variance analysis and activity-based variance analysis, the author uses a hypothetical manufacturer called Omega Technology. Omega manufactures two types of printed circuit boards, Model A and Model B. Model A is a simple, high volume product. Model B is a complex, low volume product. Annual production and sales volumes are 80,000 for Model A and 40,000 for Model B.

Since Omega is a highly automated manufacturing facility, the manufacturer has significant overhead. Controlling these indirect costs is critical to the company. Due to the level of automation, direct labor costs are negligible and are included as a part of machinery-based overhead. Therefore, total manufacturing costs consist of direct materials and factory overhead. However, the author does not discuss direct material, because it is not relevant to the ideas conveyed in the paper.

Favorable and Unfavorable Variances

In the following discussion and exhibits, the use of the terms favorable (F) and unfavorable (U) is used to classify variances. If a board requires less than the budgeted amount of activity, the variance is described as favorable (F). The terms favorable and unfavorable can be thought of in terms of the variance’s effect on net income. If a variance is favorable, the variance will have a positive effect on net income. If unfavorable, the variance will decrease net income.

Traditional Variance Overhead Analysis

As indicated in Exhibit 1, total factory overhead budget is $2,793,840 and includes the following overhead line items: Machinery, Receiving & Inspection, Setup, Material Handling, Engineering, Quality Assurance, and Facility-Level. Total overhead is broken down into variable and fixed overhead for each of these overhead cost categories.

Budgeted Overhead of Traditional Cost System (Based on Exhibit 1, p. 39)
Cost Category Variable Overhead Fixed Overhead Total Overhead
Machinery $110,000 $74,320 $184,320
Receiving and Inspection 175,000 75,000 250,000
Setup 120,000 42,240 162,240
Material Handling 90,000 31,680 121,680
Engineering 700,000 100,000 800,000
Quality Assurance 400,000 125,600 525,600
Facility Level 250,000 500,000 750,000
Total $1,845,000 $948,840 $2,793,840
Variable Overhead Rate = $1,845,000 ÷ 57,600 machine hours = $32.03125 per hour
Fixed Overhead Rate = $948,840 ÷ 57,600 machine hours = $16.47291 per hour

An assumption of traditional cost systems is that production volume drives product cost. Therefore, machine hours is used as the overhead application rate. Variable overhead is applied at a rate of $32.03125 per machine hour and fixed overhead is applied at $16.47291 per machine hour.

Actual Results

Actual units produced = Budgeted units to be produced
Actual machine hours used = 59,000
Actual Variable overhead costs = $1,845,000 (same as budgeted)
Actual Fixed overhead costs = 948,840 (same as budgeted)

Traditional Overhead Variance Analysis

The traditional variance analysis presented below is based on Exhibit 2, p. 41. The variable overhead (VO) spending variance is $44,844F and the variable overhead efficiency variance is $44,844U. These variances are calculated using machine hours as the cost driver. The favorable VO spending variance can be attributed to generating the budgeted variable overhead costs ($1,845,000) despite using extra machine hours. The unfavorable efficiency variance is due to the actual use of 59,000 machine hours instead of the budgeted 57,600 hours. The illustration also shows both fixed overhead variances (i.e., spending and production volume) are zero.

Overhead Variance Calculations - Traditional cost System

Criticisms of traditional variance analysis

Production volume is assumed to drive all overhead costs. While acceptable for external financial reporting, product costing and variance analysis are two separate functions of a cost system. Therefore, the assumptions used for product costing can be inappropriate for variance analysis purposes. When production volume is not the correct cost driver, the cost system does not reflect the true economics of production.

Managers may be led to the wrong conclusions. For example, looking at the fixed overhead variances in Exhibit 2, a manger may conclude that fixed overhead is under control (spending variance = 0) and that the plant is operating at full capacity (production volume variance = 0). Most managers will intuitively know that this is not the case. Substantial excess capacity will exist because machine hours do not have a strong relationship with variable overhead costs.

The Hierarchy of Costs in Activity-Based Costing (ABC)

Rather than assume all costs are driven by production volume, activity-based costing suggests that different manufacturing costs are associated with different types of activities. Cooper's cost hierarchy recognizes different types of activities including unit-level, batch-level, product-level, and facility-level activities.

Unit-level activities are associated with the production of a single unit. The more units produced, the more costs incurred. Therefore, production volume is the cost driver for these activities.

Batch-level activities are activities needed to produce different batches of products. In the Omega example, receiving & inspection costs, setup costs, and material handling costs are all incurred as a result of batch-level activities. Management will determine the number of batches required for a period in relation to their attitude towards holding inventory. To minimize inventory levels, larger batch sizes are used. Several small shipments increase receiving & inspection, setup, and material handling costs, while fewer, but larger shipments will decrease these costs. When a batch size equals one, unit-level and batch-level activities are indistinguishable.

Product-level activities support different products in the company’s line of products. Managers will have to consider issues such as product quality and life cycles. If a product has a short life cycle, redesigns will be necessary to keep the product marketable. This means more product-level activities.

Facility-level activities maintain a facility’s general manufacturing process. This type of activity cost can only be assigned to products arbitrarily.

Activity-Based Overhead Variance Analysis

The first step in activity-based variance analysis is to assign all overhead costs to a level of activity. Next, activity standards (standard rates) must be calculated. To reach this standard rate, the annual overhead cost is divided by the cost center’s practical capacity. Practical capacity is used so that idle capacity may be found and put to better use. The standard rates calculated for batch and product level activities do not vary with production volume. This is a fundamental difference between ABC and traditional variance analysis.

Variance analysis can be examined for the unit-level, batch-level, and product-level activities. Since facility level costs can only be assigned arbitrarily, variances cannot be calculated for this level of activity.

Exhibits 3, 4, and 5 show that under activity-based variance analysis, variances can be calculated for Model A and B boards separately. This makes it possible to discover if one product is subsidizing another. For example, in Exhibit 4 Panel B, production schedulers may be providing too much set up activity to the Model A line and not enough to the Model B line.

Unit-Level Overhead Cost Variances

By looking at the unfavorable unit-level cost variances in Exhibit 3, we can tell that there is no idle capacity. Unit-level costs are driven by production volume. Therefore, the more boards produced, the more machine hours needed. The 59,000 actual hours exceeds the standard 57,600 hours. This may tell managers to expect a greater number of machine hours in the future.

Unit Level Costs - (Adapted from Exhibit 3)
Machinery Overhead
Model Standard Quantity
(Machine Hours)
Actual Quantity
(Machine Hours)
A 38,400 39,000 $3.20 $1,920U
B 19,200 20,000 $3.20 $2,560U
Total 57,600 59,000   $4,480U
*Annual cost/Practical Capacity = $184,320 ÷ 57,600 = $3.20 per hour

Batch-Level Overhead Cost Variances

In Exhibit 4, all Model A variances are unfavorable while all Model B variances are favorable. With this pattern, management may be able to explain the reasons for the variances. In addition to the realized costs (as shown by the variances), the company may be incurring an opportunity cost with the insufficient capacity for Model B boards.

Batch-Level Costs - (Adapted from Exhibit 4 Panel A)
Receiving and Inspection
Model Standard Quantity (Shipments) Actual Quantity
A 2,100 2,350 $50 $12,500U
B 2,900 2,600 $50 $15,000F
Total 5,000 4,950   $2,500F
*Annual cost/Practical Capacity = $250,000 ÷ 5,000 shipments = $50 per shipment

Batch-Level Costs - (Adapted from Exhibit 4 Panel B)
Model Standard Quantity (Setup Hours) Actual Quantity
(Setup Hours)
A 292 350 195 $11,310U
B 540 480 195 $11,700F
Total 832 830   $390F
*Annual cost/Practical Capacity = $$162,240 ÷ 832 setup hours = $195 per setup hour

Batch-Level Costs - (Adapted from Exhibit 4 Panel C)
Material Handling
Model Standard Quantity (Material Moves) Actual Quantity
(Material Moves)
A 672 750 $65 $5,070
B 1,200 980 $65 $14,300
Total 1,872 1,730   $9,230
*Annual cost/Practical Capacity = $121,680 ÷ 1,872 material moves = $65 per move

Product-Level Overhead Cost Variances

Product-Level issues can be revealed by the variances in Exhibit 5. In Panel A, Model B requires more change orders (165) than the standard quantity (150). These extra change orders could be necessary to keep customers. The product’s life cycle may be getting shorter. Model A used only 60 percent of the standard change orders, but there is still an overall favorable variance of $80,000. This overall favorable variance tells us that although Model B is receiving more change orders, Model A is not suffering fewer change orders.

The variances related to quality assurance (Panel B) may be caused by changing customer characteristics. For example, customers of Model B may not inspect the boards before purchase, because they expect the quality level to be high, while customers of Model A inspect the product before purchase. If this is the case, Omega would not need to guarantee a high level of quality.

Product-Level Costs - (Adapted from Exhibit 5 Panel A)
Model Standard Quantity (Change Orders) Actual Quantity
(Change Orders)
A 100 60 $3,200 $128,000F
B 150 165 $3,200 $48,000U
Total 250 225   $80,000F
*Annual cost/Practical Capacity = $800,000 ÷ 250 change orders = $3,200 per change order

Product-Level Costs - (Adapted from Exhibit 5 Panel B)
Quality Assurance
Model Standard Quantity
(Test Hours)
Actual Quantity
(Test Hours)
A 55,200 40,000 $3 $45,600F
B 120,000 132,000 $3 $36,000U
Total 175,200 172,000   $9,600F
*Annual cost/Practical Capacity = $525,600 ÷ 175,200 test hours = $ 3 per test hour

Usefulness of Activity-Based Variance Analysis

Although managers are not provided with answers, activity-based variances do suggest questions that managers will need to look into. Manager’s attention can be directed toward specific areas or product lines that need attention to improve profitability. For instance, activities with idle capacity can be identified and corrected. “Troublesome” products or customers can be identified. The variance analysis in Exhibit 5 suggests that customers who demand excessive engineering or product changes, may be too costly to maintain. Ways of addressing variances are suggested by the analysis. In other words, if there is a variance in a product-level activity, then a product-level activity should be investigated, not a unit or batch-level activity.


Related summaries:

Cheatham, C. B. and L. R. Cheatham. 1996. Redesigning cost systems: Is standard costing obsolete? Accounting Horizons (December): 23-31. (Summary).

Cooper, R. and R. S. Kaplan. 1992. Activity-based systems: Measuring the costs of resource usage. Accounting Horizons (September): 1-13. (Summary).

Martin, J. R. Not dated. Chapter 4: Normal Historical Full Absorption Job Order Costing. Management Accounting: Concepts, Techniques & Controversial Issues. Management And Accounting Web.

Martin, J. R. Not dated. Chapter 7: Activity Based Product Costing. Management Accounting: Concepts, Techniques & Controversial Issues. Management And Accounting Web.

Martin, J. R. Not dated. Chapter 10: Standard Full Absorption Costing. Management Accounting: Concepts, Techniques & Controversial Issues. Management And Accounting Web.

Martin, J. R. Not dated. Chapter 13: Profit Analysis: An Overall Performance Evaluation - Part I.  Management Accounting: Concepts, Techniques & Controversial Issues. Management And Accounting Web.

Martin, J. R. 2000. The advantages of teaching three production volume variances. Journal of Accounting Education 18(1): 35-50. Example of Variance analysis in an ABC system.

Stammerjohan, W. W. 2001. Better information through the marriage of ABC and traditional standard costing techniques. Management Accounting Quarterly (Fall): 15-21. (Summary).

Wing, K. T. 2000. Using enhanced cost models in variance analysis for better control and decision making. Management Accounting Quarterly (Winter): 27-35. (Summary).